BRTCrowd

BRTCrowd is a mobile + web AI app that predicts bus rapid transit crowding by stop and by vehicle, so riders can choose less-packed departures and agencies can spot overload patterns. It combines real-time GTFS-Realtime feeds with lightweight rider check-ins (one-tap “packed/ok/empty”), optional passive signals (accelerometer-based dwell-time inference), and historical patterns to estimate occupancy and platform congestion. The rider experience is simple: open the map, see the next arrivals with a crowding score, and get alerts when your usual trip is likely to be uncomfortably full. For agencies, a web dashboard highlights chronic pinch points (specific stops, times, and routes) and quantifies how often buses leave riders behind. Brutal truth: this only works if you secure data access and reach enough rider participation in a corridor; without that, predictions degrade and churn rises.